Haight, R., 2002. An Inventory and Assessment of the Finney Creek Riparian Forest. Skagit System Cooperative, La Conner, WA. pp. 59.

Inventories of riparian area vegetation are a valuable component to riparian forest management. This is a simple statement and indeed seems nearly redundant. However, relatively few comprehensive inventories of riparian areas have been published. Likewise, few methodologies have been developed to produce such inventories. The question in all forest inventories is one of how much, but in riparian systems an additional question is where. That is, where are components of the riparian systems changing with regard to riparian function, where are functions impeded, and what positions do riparian vegetation elements hold with respect to the channel? These are some of the possible questions one would consider and develop data collection procedures for in a riparian system.
Methodologies have been developed for broad scale generalizations of riparian systems, such as in watershed analysis (Washington Forest Practices Board 1995). In these analyses, riparian vegetation might be classified into general categories such as “mature hardwood” or “young conifer”. Methods such as this might focus on channel adjacent vegetation as was a quick way of generalizing at the watershed scale. Other methods have focused on desired future conditions or old growth structure. For example, a desired future condition methodology might use randomly located plots of given criteria and collect measurements on riparian stands of mature ages. Measured amounts might then be used as targets for developing stand management guidelines in riparian zones. Yet others have used randomly selected sites to characterize riparian zones in an aggregate fashion. These might measure several attributes of overstory and understory vegetation across sites located in different stream networks and extrapolate the findings to a study area.
One common thread in most published studies of riparian systems is the use of a randomized approach to sampling. Random approaches have long been the preferred method by many wishing to characterize populations of various types. Random approaches are discussed by hundreds of authors, see Platts et al. (1987) or Husch et al. (1982), or any standard mensuration text. The basic premise behind utilizing random sampling approaches are that results can be